Single ended EL34 amp design

[bepone]

On the heater side, we are both looking at a 3A supply. Not enough to feed a regulated supply to all the tubes perhaps (the EL34 is hungry) but possibly enough to provide a regulated supply for the preamp tubes and AC for the power tube.

only first tube... but anode and g2 of the output must be clean also

[Stephen1966]

I would like to ask you - how do you implement a Presence control in a SE cathode biased amp? And, do you need a NFB net for this to work?

when you are playing with negative feedback resistor, this is already some kind of presence because is affecting the highs, but i saw somewhere schematic with this function on SE amp..
but when the amp is setup good, it is not needed

True, but I ask because a presence control may be supplemented with a resonance control and for my harp player this would be the bees knees. Simply, if we can fit a presence control, we can also fit a resonance control and there is no reason why both could not be fitted.

I am also looking into how we integrate these filters in a SE amp and if I find anything, I will post it. Do you recall where you found the SE design for presence? It may not be that difficult - coming off a cathode but Merlin leaves me with the impression it can only work in conjunction with NFB.

(Sorry if I have misunderstood you, Merlin!)

[Stephen1966]

On the heater side, we are both looking at a 3A supply. Not enough to feed a regulated supply to all the tubes perhaps (the EL34 is hungry) but possibly enough to provide a regulated supply for the preamp tubes and AC for the power tube.

only first tube... but anode and g2 of the output must be clean also

I don't think the problem will be so great at the power stage - a little hum can be expected, and in a studio situation those frequencies can be filtered out. Still, it would be nice to run all the tubes with a regulated supply off a 3A coil. The power factor loss of a standard DC circuit is too great but Merlin's switching supply with a voltage regulator looks promising and if we can get a soft start up as well, not stressing the heaters with any voltage/current spikes, even better.

As an object lesson, the Champ resissues have terrible noise under certain conditions and these are not regulated. I am confident however, that with the right circuit it is possible to feed a constant DC supply to the heaters for several preamp tubes and the power tube all from a 3A coil. It is already pretty close to the required capacity and just obtaining a little more efficiency should be possible without dropouts where we exceed the capabilities of the transformer. The transformer may be able to supply greater current demands for short periods of time but then it starts to overheat.

[Stephen1966]

Yeah! That's kinda what I'm thinking. Interesting PT choice - 100mA off the HT - is that adequate enough to drive the plates under peak conditions? I'm still not decided between:

yes ma this 100mA is working only half cycle, so at least double current is possible to extract (and this is not needed ) it is A class, so current will be around 80mA permanent draw

Perhaps you can elaborate what you mean by "half cycle"? I don't recall ever reading about this. A general consensus is that manufacturers underrate their components for safe operation, but is 'half-cycle' a thing peculiar to transformers? Elsewhere, "80mA permanent draw" looks like your quiescent currents for a 12AX7 + 12AT7 + EL34 with your selected operating point. Do I understand you correctly?

[bepone]

I don't think the problem will be so great at the power stage - a little hum can be expected, and in a studio situation those frequencies can be filtered out.

exactly in the power stage IS A problem, because Single Ended topology has one element in the output , PP has two, and there is no power supply humm cancelation like in PP.. some of the humm is passed to the output-speaker, less if g2 is nicely regulated.

Thats why anode must be connected to second el.capacitor, so CLC supply is the best, C1, L, C2, anode feed is from C2

[bepone]

Perhaps you can elaborate what you mean by "half cycle"? I don't recall ever reading about this. A general consensus is that manufacturers underrate their components for safe operation, but is 'half-cycle' a thing peculiar to transformers? Elsewhere, "80mA permanent draw" looks like your quiescent currents for a 12AX7 + 12AT7 + EL34 with your selected operating point. Do I understand you correctly?

Every 250VAC secodary is working only half of sinusoide, rectifiying only half of time, other half of cicle is switched off.

Trafo 250V-0-250V 100mA, double secondary, rectified with 2 diodes, is equal to 500VAC CT, end-to-end, 100mA, or 50VA of supplied power, speaking briefly.

It is the same like single secodary 250VAC trafo with 200mA capability, rectified with 4 diodes, or same 50VA of power capability.

But is not so simple calculation, from 250V i will get in best case close to 350V.. and from winding of 50VA without excessive heating is possible to extract cca 50/350=140mA * efficiency of transformer.. maybe 80% or less.. so we will come to 110mA @ 350V, and also little bit less voltage, so 110mA*340VDC from mentioned trafo.. 80mA will be fine

[Stephen1966]

Perhaps you can elaborate what you mean by "half cycle"? I don't recall ever reading about this. A general consensus is that manufacturers underrate their components for safe operation, but is 'half-cycle' a thing peculiar to transformers? Elsewhere, "80mA permanent draw" looks like your quiescent currents for a 12AX7 + 12AT7 + EL34 with your selected operating point. Do I understand you correctly?

Every 250VAC secodary is working only half of sinusoide, rectifiying only half of time, other half of cicle is switched off.

Trafo 250V-0-250V 100mA, double secondary, rectified with 2 diodes, is equal to 500VAC CT, end-to-end, 100mA, or 50VA of supplied power, speaking briefly.

It is the same like single secodary 250VAC trafo with 200mA capability, rectified with 4 diodes, or same 50VA of power capability.

But is not so simple calculation, from 250V i will get in best case close to 350V.. and from winding of 50VA without excessive heating is possible to extract cca 50/350=140mA * efficiency of transformer.. maybe 80% or less.. so we will come to 110mA @ 350V, and also little bit less voltage, so 110mA*340VDC from mentioned trafo.. 80mA will be fine

That's useful to know, thank you. So in other words you get the same power handling from a centre-tapped HT secondary and half-wave bridge, as you do from a non-centre-tapped HT secondary with a full wave bridge?

[Stephen1966]

Perhaps you can elaborate what you mean by "half cycle"? I don't recall ever reading about this. A general consensus is that manufacturers underrate their components for safe operation, but is 'half-cycle' a thing peculiar to transformers? Elsewhere, "80mA permanent draw" looks like your quiescent currents for a 12AX7 + 12AT7 + EL34 with your selected operating point. Do I understand you correctly?

Every 250VAC secodary is working only half of sinusoide, rectifiying only half of time, other half of cicle is switched off.

Trafo 250V-0-250V 100mA, double secondary, rectified with 2 diodes, is equal to 500VAC CT, end-to-end, 100mA, or 50VA of supplied power, speaking briefly.

It is the same like single secodary 250VAC trafo with 200mA capability, rectified with 4 diodes, or same 50VA of power capability.

But is not so simple calculation, from 250V i will get in best case close to 350V.. and from winding of 50VA without excessive heating is possible to extract cca 50/350=140mA * efficiency of transformer.. maybe 80% or less.. so we will come to 110mA @ 350V, and also little bit less voltage, so 110mA*340VDC from mentioned trafo.. 80mA will be fine

That's useful to know, thank you. So in other words you get the same power handling from a centre-tapped HT secondary and half-wave bridge, as you do from a non-centre-tapped HT secondary with a full wave bridge?

In any case, I'm going with a solid-state rectifier here. There are for me at least, no obvious advantages from a tube rectifier. Still, to repeat, I don't ever recall anyone ever talking about this so this is really useful to know. I'm now thinking that I've read of a way to determine the VA of a transformer when it isn't included in the datasheets - I need to go back to my books - but from what you are saying (which sounds good to me) we can estimate the efficiency of the transformer at 80%.

The CLC filter with power tube anode connecting to the second cap also sounds like a good approach. I didn't think it would make such a difference given the higher signal level there but I will defer to your better understanding of the SE power output properties. It makes sense - what you say about the difference between SE and PP topologies. Really, I'm willing to try anything to tame the power frequency when SE amps have a notorious reputation for loud hum. As long as there is a screen stopper on g2 there appears no reason why the top of the resistor can't operate at the same voltage as the anode.

[Colossal]

Stephen,

I think for SE amps, a CLC in front of B+1 is essential. It really lowers the noise floor. Some years ago, I built an SE amp with a 20W OT with three different primary taps so I could try any power tube and be able to switch primaries and it was absolutely quiet at idle. Very soft hiss, but no hum whatsoever. As for a tube rectifier, they can be useful for where you might want to operate the amp at a lower voltage for one power tube(s) or higher for another, or just see what sounds and feels best. But for convenience and reliability and one less tube to worry about, SS is simple and easy.

[Stephen1966]

Stephen,

I think for SE amps, a CLC in front of B+1 is essential. It really lowers the noise floor. Some years ago, I built an SE amp with a 20W OT with three different primary taps so I could try any power tube and be able to switch primaries and it was absolutely quiet at idle. Very soft hiss, but no hum whatsoever. As for a tube rectifier, they can be useful for where you might want to operate the amp at a lower voltage for one power tube(s) or higher for another, or just see what sounds and feels best. But for convenience and reliability and one less tube to worry about, SS is simple and easy.

Got it! Brought up on PP topologies this seems less of a concern with their hum-rejection properties. Easy to forget this when diving into SE design. I see it is absolutely do-able with CLC before the anode though so I'm not even going to try to second guess you. Just need to figure out the values...

C = I / (2 f Vripple)

With I = DC average load, f = mains frequency, B+ = 350V (in my design) and Vripple at 5% or less (350 x 0.05 = 17.5V)

I can insert the values from my design into this and following Merlin https://www.valvewizard.co.uk/smoothing.html along with some simulation, I can piece this together.

[Stephen1966]

Just as a means of understanding the design choices better, I transferred all of Kirk Elliot's Remaking Champ schematics into a single file.

Remaking Champ - schematic.pdf

Interestingly, instead of a CLC net before the power tube anode, we have a CRC net here.

I'm not likely to use much (if any) of this, but I thought it might be useful to see the circuit as a whole and I might run up a simulation of this in Spice to investigate the filtering effects Elliot was aiming for.

[Littlewyan]

A CRC can be used instead of a CLC. It’s cheaper to implement than a CLC and still better than having just C. All depends on your budget and space constraints. The downsides are you could end up having to either use larger capacitors or a higher resistance to get an equal filter to a CLC and thus dropping more voltage in the process.

If you’ve ever replaced a choke in a Marshall with a 100Ohm resistor you’d notice the noise floor increases.

[Stephen1966]

A CRC can be used instead of a CLC. It’s cheaper to implement than a CLC and still better than having just C. All depends on your budget and space constraints. The downsides are you could end up having to either use larger capacitors or a higher resistance to get an equal filter to a CLC and thus dropping more voltage in the process.

If you’ve ever replaced a choke in a Marshall with a 100Ohm resistor you’d notice the noise floor increases.

Good points. I think when you consider the effect of CRC on the noise floor all the other disadvantages (cost, space and voltage drop) melt away. Money and space can all be found (or lost) and a lower voltage can be compensated for. But there isn't much you can do, outside of a recording studio, with a noisy sound floor. It would follow you around wherever you go.

I don't recall the chassis size I already have, it's about twice the area of a Champ chassis though and my early drawings account for the inclusion of a choke so it's a CLC network I am planning. The general consensus seems to be that SE class A can be quite noisy so a design consideration here is to build noise suppression in. Anything I can do as long as it doesn't detract from the tone and character of the design.

[rogb]

You are probably right, I like this transformer because it looks well built and has a good bandwidth compared to similar Hammond transformers for example. I'm not even sure I will stay with this though, there are possibilities for extra gain using different primary impedances.

What do you think of the power transformer though?

In chosing the SE OT's, DC current is No1 factor! And must be optimised, exactly for this design! Why?

Because in SE OPT you have DC and AC magnetisation in the core, DC magnetisatiion is more important, to "bias" the core exactly in the linear part of the HB curve! Somewhere in the half (this is manufacturer problem if is not like this, and is difficult to measure at home).

We dont want to take 200mA SE trafo if we need only 150! If do, then with 150mA we will bias DC magnetic flux through the core in the sub optimum area, and saturation and distortion of the transformer will limit the output power! We want flux to be in the half of linear area..

The same for output power, no need to take 20W if we need 10W, we will just have lower inductance L, because with bigger core we take lower number of the turns..
Bad for our bass side. Smaller core with higher number of the turns is better for guitar, and tightness on the bass side..

SE OT still must be bigger than PP transformer for the same power because is wasting half of the flux for nothing, only to bias output tube.

So some medium transformer will pass nicely, not too small, not too big

This OPT https://primarywindings.com/product/500 ... ansformer/ has 100mA DC current at the primary - any good? Quiescent Ia = 63mA and the theoretical peak - point (A) on the chart is 133mA. I would expect some heat to be generated but the tube is unlikely to ever reach that peak. It would take some really heavy sustained overdrive to get it there!

The 20W output is overkill but regarding that bandwidth, if the manufacturers claims are true it has a much better low end response than the Hammond. A 15Hz roll-off versus a 100Hz roll-off if I remember it well.

Hi I just built an SE EL34 5F2A with this 15w OT from the same firm.
https://primarywindings.com/product/amp ... ansformer/

[Stephen1966]

In chosing the SE OT's, DC current is No1 factor! And must be optimised, exactly for this design! Why?

Because in SE OPT you have DC and AC magnetisation in the core, DC magnetisatiion is more important, to "bias" the core exactly in the linear part of the HB curve! Somewhere in the half (this is manufacturer problem if is not like this, and is difficult to measure at home).

We dont want to take 200mA SE trafo if we need only 150! If do, then with 150mA we will bias DC magnetic flux through the core in the sub optimum area, and saturation and distortion of the transformer will limit the output power! We want flux to be in the half of linear area..

The same for output power, no need to take 20W if we need 10W, we will just have lower inductance L, because with bigger core we take lower number of the turns..
Bad for our bass side. Smaller core with higher number of the turns is better for guitar, and tightness on the bass side..

SE OT still must be bigger than PP transformer for the same power because is wasting half of the flux for nothing, only to bias output tube.

So some medium transformer will pass nicely, not too small, not too big

This OPT https://primarywindings.com/product/500 ... ansformer/ has 100mA DC current at the primary - any good? Quiescent Ia = 63mA and the theoretical peak - point (A) on the chart is 133mA. I would expect some heat to be generated but the tube is unlikely to ever reach that peak. It would take some really heavy sustained overdrive to get it there!

The 20W output is overkill but regarding that bandwidth, if the manufacturers claims are true it has a much better low end response than the Hammond. A 15Hz roll-off versus a 100Hz roll-off if I remember it well.

Hi I just built an SE EL34 5F2A with this 15w OT from the same firm.
https://primarywindings.com/product/amp ... ansformer/

What's your experience there? I heard about them through Merlin's website. They look good on paper at least!